Synthesis of β-pyrrolic-modified porphyrins and their incorporation into DNA

Article abstract of DOI:10.1002/chem.201003200

A synthetic methodology for the synthesis of various β-pyrrolic- functionalised porphyrins and their covalent attachment to 2′-deoxyuridine and DNA is described. Palladium(0)-catalysed Sonogashira and copper(I)-catalysed Huisgen 1,3-dipolar cycloaddition

Full text of DOI:10.1002/chem.201003200

FULL PAPER
DOI: 10.1002/chem.201003200
Synthesis of b-Pyrrolic-Modified Porphyrins and Their
Incorporation into DNA
Adam W. I. Stephenson,[a] Ashton C. Partridge,[a, b] and Vyacheslav V. Filichev*[a]
Abstract: A synthetic methodology for
the synthesis of various b-pyrrolic-func-
tionalised porphyrins and their cova-
lent attachment to 2’-deoxyuridine and
DNA is described. Palladium(0)-cata-
lysed Sonogashira and copper(I)-cata-
lysed Huisgen 1,3-dipolar cycloaddition
reactions were used to insert porphy-
temperature-dependent i-motif struc-
tures. Porphyrin insertion also led to
the aggregation of single-stranded
purine–pyrimidine sequences, which
could be dissociated by heating at
908C for 5 min. Parallel triplexes and
anti-parallel duplexes were formed in
the presence of the appropriate com-
plementary strand(s). Depending on
the modification, porphyrins were
placed in the major and minor grooves
of duplexes and were used as bulged
intercalating insertions in duplexes and
triplexes. In general, the thermal stabi-
lisation of parallel triplexes possessing
porphyrin-modified triplex-forming oli-
gonucleotide (TFO) strands was ob-
served, whereas anti-parallel duplexes
were destabilised. These results are
compared and discussed on the basis of
the results of molecular modelling cal-
culations.
ACHTUNGTRENNUNGrins into the structure of 2’-deoxyuri-
dine and DNA. Insertion of a porphy-
rin into the middle of single-stranded
CT oligonucleotides possessing a 5’-ter-
minal run of four cytosines was shown
to trigger the formation of pH- and
Keywords: click chemistry · cyclo-
addition · DNA · porphyrinoids ·
thermal stability
Introduction
furanose residues,[10–15] the phosphate backbone[16,17] and acy-
clic linkers.[18,19] Porphyrin moieties have been introduced as
3’- or 5’-molecular caps,[13,19–21] as a nucleobase substituent in
the middle of the helix[18] or as a label in the minor[11,17,22]
and major[5–9] grooves. Several tethers have been used for
the covalent attachment of porphyrins including maleimido-
thiols,[8,17] amides,[11–13,15] phosphates[16] and alkynes.[4–7,9] The
incorporation of the porphyrins as 5ꢀ and 3ꢀ caps has been
shown to improve base-pair fidelity and duplex stability[10]
as well as showing the ability for the porphyrins to electroni-
cally communicate with each other (exciton coupling). This
feature enables porphyrins to be used as detectors of DNA
secondary structures in circular dichroism (CD) spectrosco-
py.[12–16,21] The internal modification of ONs with porphyrins
has recently become of interest for the development of heli-
cal scaffolds. Such examples include the synthesis of a DNA
containing 11 meso-functionalised porphyrins on a single
strand[5,6] and a zipper porphyrin assembly in the major
groove of a DNA duplex.[7]
DNA is an attractive supramolecular scaffold for the pro-
duction of functionalised arrays due to its favourable charac-
teristics such as its ability to functionalise DNA building
blocks at will, the spontaneous self-assembly of complemen-
tary strands and the formation of helices as well as 2D and
3D structures.[1] It can be recognised by proteins and other
nucleic acid binding compounds. The covalent labelling of
DNA, which results in modified oligonucleotides (ONs), can
take place through the phosphate backbone, the nucleobase
or the ribose sugar, locating the desired molecule in various
positions around the DNA double helix, which allows for
the production of controlled helical arrays. Porphyrins are
useful labels and have been widely studied in a variety of
applications including in light-harvesting devices[2] and the
production of reactive oxygen species.[3]
The site-specific covalent attachment of porphyrin moiet-
ies to DNA has been achieved by using a variety of method-
ologies including the modification of nucleobases,[4–9] ribo-
In contrast to the common functionalisation of a porphy-
rin at the meso position,[5–9,11,20] which results in a system
nearly orthogonal to the porphyrin core, we have recently
synthesised a b-pyrrolic-modified porphyrin and shown that
its H-aggregate formation in the minor groove significantly
stabilised duplexes (DTm per modification=+7.5–7.98C).[22]
Duplex thermal stability exceeded 908C when four porphy-
[a] Dr. A. W. I. Stephenson, Prof. A. C. Partridge, Dr. V. V. Filichev
College of Sciences, Institute of Fundamental Sciences
Massey University, Private Bag 11-222
Palmerston North (New Zealand)
Fax : (+64)6-350-5682
E-mail: v.filichev@massey.ac.nz
AHCTUNGTREGrNNNU ins were stacked in the zipper motif as a result of superior
[b] Prof. A. C. Partridge
interactions between b-pyrrolic-modified porphyrins in com-
parison with previously studied DNA–porphyrin con-
structs.[5–8,17] We were curious, therefore, as to the effect of
different attachments of b-pyrrolic porphyrin to DNA on
Present address: Auckland University, 20 Symonds Street
Auckland (New Zealand)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201003200.
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6227

binding affinity and the spectroscopic properties of DNA–
porphyrin conjugates.
Results and Discussion
In this manuscript we report the synthesis of b-pyrrolic-
functionalised porphyrins covalently attached to uridines
using Sonogashira or copper(I) azide/alkyne cycloaddition
(CuAAC)[23] reactions. To screen the effect of different por-
phyrin locations in DNA duplexes, we preferred a post-syn-
thetic modification of DNA[24] using CuAAC in which a
functional group of the label (azide) reacts with the comple-
mentary functional group on the DNA (terminal alkyne).[25]
This approach is more convenient than the time-consuming
preparation of individual phosphoramidites especially taking
into account the limited lifetime of porphyrin-containing
phosphoramidites.[11,12]
Apart from our recent publication,[22] there has only been
one other report of b-pyrrolic-substituted porphyrin in the
structure of DNA.[18] A flexible tether and phosphoramidite
chemistry were used to place a porphyrin in the middle of a
DNA duplex as a base substituent. We decided to investi-
gate the possibilities of porphyrin covalent attachment to
nucleosides through the b-pyrrolic position with the ultimate
goal of creating DNA-based porphyrinic assemblies. To
form a covalent linkage with nucleosides/DNA we consid-
ered two options that can be applied to either pre- or post-
synthetic modification of DNA, that is, palladium(0)-cata-
lysed Sonogashira and copper(I)-catalysed azide/alkyne cy-
cloaddition.[24] The synthesis of porphyrin precursors suita-
ble for coupling reactions on nucleosides and DNA is de-
scribed below and is based on the 2-formyl-5,10,15,20-tetra-
phenylporphyrin 10 (Scheme 2).
We prepared ONs possessing insertions of 2’-deoxy-5-
ethynyluridine[26] (V; Scheme 1), 2’-O-propargyluridine[27]
(Y) or (R)-4-ethynylphenylmethylglycerol[28] (Z) and used
This compound is easy to syn-
thesise in high yields and multi-
gram quantities without the
need for difficult silica gel chro-
matography. This is in contrast
to another building block used
in b-pyrrolic functionalisation,
2-bromo-5,10,15,20-tetraphenyl-
porphyrin. The radical bromina-
tion of 5,10,15,20-tetraphenyl-
porphyrin (TPP) using N-bro-
mosuccinimide (NBS) in organ-
[29]
ic solvents such as CHCl3
usually produces di- and tribro-
minated species along with the
target compound, which is diffi-
cult to purify particularly
during scaling up. The aldehyde
in 10 serves as a useful func-
tionality for the synthesis of dif-
ferent alkene and alkyne deriv-
atives possessing
a range of
functional groups, which in turn
can be used for covalent attach-
ment to DNA. We assumed
Scheme 1. Structures of porphyrin-functionalised ONs produced by CuAAC.
that variation of the porphyrin
aromatic (n=0) or aliphatic (n=1) azidoporphyrins in sub-
sequent conjugation reactions. This allowed the placement
of porphyrins in major (1 and 2) and minor grooves (3 and
4) of the duplex as well as the use of porphyrins as interca-
lating bulged insertions in duplexes and triplexes (5 and 6).
Thermal stability and circular dichroism studies were under-
taken to assess the effect of each porphyrin modification on
single-stranded, duplex and triplex DNA.
attachment and positioning within the DNA structures is of
interest due to the increased attention shown in recent years
in macromolecular multi-chromophoric scaffolding in the
field of material sciences.[30]
Synthesis of porphyrin-2’-deoxyuridine conjugates using the
Sonogashira reaction: The b-functionalised porphyrin nu-
cleoside 18 was synthesised in five steps from 2-formyl-
5,10,15,20-tetraphenylporphyrin 10 (Scheme 2, see the Sup-
porting Information for experimental procedures). Com-
pound 10 was easily prepared according to the procedure of
Bonfantini et al.[31] in 85% overall yield in three steps by
the quantitative insertion of copper(II) into TPP followed
6228
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238

b-Pyrrolic-Modified Porphyrins
FULL PAPER
Scheme 2. Reagents and conditions: i) diphenyl phosphite, MgO, RT, overnight, 90%; ii) CH2Cl2, DDQ, PPh3, nBu4NBr, RT, overnight, 84%; iii) THF,
tBuOK, RT, 3 h, 75%; iv) CHCl3, MeOH, Zn(OAc)2·2H2O, RT, 1 h, 99%; v) Et3N, trimethylsilylacetylene, [Pd(PPh3)4], CuI, reflux, 3 h, 93%; vi) CH2Cl2,
THF, TBAF, RT, 5 min, 97%; vii) Et3N, [Pd(PPh3)4], CuI, 708C, overnight, 78% crude, 12% pure; viii) CH2Cl2, TFA, RT, 2 min, 70%.
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
by
Vilsmeier
formylation
and
demetallation.
protected porphyrin 13 repeatedly with 5% aq. Na2EDTA
solution followed by 3m NH4OH. Failure to do so resulted
in the exclusive formation of the Glaser homodimer byprod-
uct 14 upon cleavage of the silyl protecting group. After
washing, deprotection of 13 to 15 using TBAF was achieved
with no sign of the unwanted Glaser homodimer 14.
The target nucleoside 18 (Scheme 2) was synthesised from
2-(4’-ethynylphenyl)ethynyl-5,10,15,20-tetraphenylporphyrin-
AHCTUNGTREGaNNUN tozinc(II) (15) and 2’-deoxy-5’-O-DMT-5-iodouridine (17)
(DMT=4,4’-dimethoxytrityl) in high yield. Crucial to the
success of the Sonogashira coupling reaction between 15
and 17 was the complete degassing of all the solvents and
the use of at least 4 equiv of 17 to maximise the formation
of 18. It was also essential to mix both reactants in Et3N
before the addition of the palladium(0) and copper(I) cata-
lysts. Failure to do so resulted in the formation of the homo-
dimer 14. Although the synthesis was successful, purification
by silica gel chromatography was problematic and pure
compound 18 was obtained in only 12% yield after subse-
quent methanol precipitation. Attempts to couple iodopor-
phyrin (12) and 2’-deoxy-5’-O-DMT-5-ethynyluridine (16)
by using the same reaction conditions showed only trace
BromophosphoACHTUNGTRENNUNG
nate 9, required for the synthesis of 11, was
obtained from phosphonate 8 by using tetrabutylammonium
bromide, DDQ and triphenylphosphine in 84% yield by
using the method of Firouzabadi et al.[32] Phosphonate 8 was
synthesised in 90% yield by using diphenyl phosphite and 4-
iodobenzaldehyde (7). The iodo-functionalised porphyrin 11
was obtained by a modified Horner–Emmons reaction[33] of
the bromophosphonate 9 and TPP-CHO (10) in 75% yield.
This reaction could easily be scaled up to allow the multi-
gram production of 11. Zinc(II) was inserted into the por-
phyrin core to give 12 in quantitative yield to prevent
copper insertion during the subsequent Sonogashira reac-
tions in which copper(I) is used as a co-catalyst. Conversion
of 12 into 13 was achieved in 93% yield by heating at reflux
a
solution
tetraphenylporphyrin
lene in Et3N in the presence of 0.3 equiv of [PdACHTUNGTRENNUNG
of
2-(4’-iodophenyl)ethynyl-5,10,15,20-
atozinc(II) (12) and trimethylsilylacety-
(PPh3)4] and
AHCTUNGTRENNUNG
0.5 equiv of CuI under argon overnight. The use of DMF as
solvent resulted in the isolation of the starting material. It
was critical to remove trace quantities of copper and palladi-
um salts by washing the CH2Cl2 solution of trimethylsilyl-
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6229

V. V. Filichev et al.
amounts of the desired product by TLC analysis of the reac-
tion mixture. Demetallation and DMT deprotection of 18
performed in one step by using trifluoroacetic acid provided
compound 19, the spectroscopic properties of which were
compared with the previously reported meso-linked uridine
nucleoside 20.[5] In chloroform, the B band of the meso-uri-
dine 20 was observed at 420 nm whereas that of the b-pyr-
rolic-functionalised 19 was detected at 429.5 nm. This 9.5 nm
bathochromic shift in the B band of 19 is most likely a
result of a higher degree of conjugation between the por-
phyrin core and uracil in comparison with 20. To the best of
our knowledge, this is the first reported synthesis of a b-pyr-
rolic porphyrin linked to a nucleobase. Owing to the unex-
pectedly low quantities of 18 obtained, conversion of pure
18 into the corresponding phosphoramidite was not attempt-
ed and efforts were focused on CuAAC and post-synthetic
DNA modifications.
azide 26. Azides 24 and 25 were synthesised from the phos-
phonium salt 27, which was obtained from TPP in six steps
in an overall yield of 85% via the formation of TPP-CHO
(10).[34] Treatment of a CH2Cl2 solution of 27 with 4-
nitrobenzACTHUNRTGENNUGaldehyde and 1,8-diazabicycloHCATUNGTRENN[UGN 5.4.0]undec-7-ene
(DBU) using a modification of the method of Bonfantini
et al.[31] resulted in the rapid formation of a cis/trans mixture
of 28, which was converted exclusively into the trans isomer
by treatment with iodine in CHCl3. Metallation of the nitro
derivative 28 with NiACHTNUTRGNEUNG(OAc)2·4H2O provided the nickel(II)
analogue, which was reduced to the amine 29 by using
SnCl2/HCl. Nickel(II) was inserted into the core of the por-
phyrin before the reduction of the nitro group to prevent in-
sertion of tin during the reduction reactions and copper
during subsequent CuAAC reactions. The b-functionalised
aromatic azide 24 was prepared using
a
modified
method[35,36] by diazotisation with H2SO4/NaNO2 in darkness
followed by the addition of NaN3.
Synthesis of porphyrin-2’-deoxyuridine conjugates using
CuAAC: Three different approaches were used to prepare
the 1,4-regioisomeric 1,2,3-triazole-linked porphyrin nucleo-
sides 21–23 (Scheme 3) as models for the post-synthetic
CuAAC reactions. To demonstrate the versatility of the re-
action three different porphyrin azides were chosen: the
alkene-linked aromatic azide 24, the more flexible alkene-
linked aliphatic azide 25 and the acetylene-linked aromatic
The aliphatic azide 25[22] was synthesised by a Wittig reac-
tion between phosphonium salt 27 and 4-(azidomethyl)benz-
AHCTUNGTRENNUNG
aldehyde[37] in the presence of DBU followed by isomerisa-
tion with iodine and nickel(II) insertion. The azide 26 was
synthesised in 61% yield by the reaction of the iodo precur-
sor 12 with NaN3, sodium ascorbate, N,N-DMEA and Cu-
AHCTUNTGRENGUN(N ACN)4PF6 in dry DMSO. Alternative reaction conditions
using CuI as catalyst or toluene as solvent resulted in the
isolation of the starting material
12.
The model triazoles 21–23
were synthesised from the cor-
responding azides 24–26 and 2’-
deoxy-5’-O-DMT-5-ethynyluri-
dine (16) with 2 equiv of Cu-
AHCTNUTGREGN(NNU ACN)4PF6 in THF. This copper
catalyst, which is soluble in or-
ganic solvents, was used as it
had previously shown improved
results in comparison with
CuSO4·5H2O or CuI in CuAAC
reactions.[35] Silica gel TLC
analysis of the reaction mixture
showed the formation of
a
more polar material that moved
either as one or two spots in
MeOH/CH2Cl2 (1:19). This sug-
gested the complete (21) or par-
tial cleavage (22 and 23) of the
DMT protecting group during
the CuAAC reaction, which
was confirmed by 1H NMR
spectroscopy and ESI-MS anal-
ysis of the intractable mixture.
The desired triazoles were
easily purified from the un-
reacted azides by silica gel
column chromatography. As ex-
pected, reactions performed
Scheme 3. Reagents and conditions: i) 4-nitrobenzaldehyde, DBU, CH2Cl2, RT, 30 min, then CHCl3, I2, RT, 3 h,
85%; ii) CHCl3, MeOH, Ni
78%; iv) THF, H2O, NaNO2, H2SO4, RT, 2 h then NaN3, RT, 20 min, 98%; v) 4-(azidomethyl)benzaldehyde,
DBU, CH2Cl2, RT, 20 min, then CHCl3, I2, RT, overnight, 60%; vi) CHCl3, MeOH, Ni(OAc)2·4H2O, reflux,
overnight, 99%; vii) DMSO, NaN3, sodium ascorbate, Cu(ACN)4PF6, N,N-DMEA, 708C, 48 h, 61%; viii) THF,
Cu(ACN)4PF6, RT, 2–4 days, 36% for 21, 28% for 22 and 39% for 23.
ACHTNUGTRNEUN(GN OAc)2·4H2O, reflux, overnight, quantitative; iii) THF, SnCl2·2H2O, HCl, RT, 48 h,
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
6230
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238

b-Pyrrolic-Modified Porphyrins
FULL PAPER
with the free base azide 30 resulted in the quantitative isola-
tion of the copper(II)-metallated porphyrinic azides with no
triazole formation.
ACHTUNGTRENNUNG
The spectroscopic properties of the porphyrin-linked
uracil derivatives 21–23 were investigated and compared. In
the UV/Vis spectra of compound 23 the porphyrin B band
occurred at 435 nm. By comparison with the ethynyl deriva-
tive 18, which has a B band at 440 nm, we can conclude that
the introduction of the triazole in 23 leads to a loss of conju-
gation between uracil and the porphyrin. Similarly, a com-
parison of the aromatic triazole 21 with aliphatic triazole 22,
which show B bands at 427.5 and 426 nm, respectively, con-
firmed that the incorporation of CH2 results in disruption of
the remaining conjugation between the porphyrin core and
the uracil.
Post-synthetic DNA modifications using the Sonogashira
reaction: Post-synthetic oligonucleotide modification is an
attractive alternative to the time-consuming preparation of
multiple phosphoramidites. This is particularly important for
screening various positions of the porphyrin moiety in the
structure of DNA. In addition, due to the limited stability of
porphyrin-containing phosphoramidites[11,12] and difficulties
in obtaining the pre-synthetic analogues in sufficient quanti-
ties, as discussed above, post-synthetic modification is clear-
ly an option for porphyrin–DNA derivatives. Therefore we
decided to investigate post-synthetic Sonogashira and
CuAAC reactions as a means of creating porphyrin-modi-
fied ONs.
Post-synthetic DNA modifications using CuAAC: The cop-
per(I)-catalysed Huisgen 1,3-dipolar cycloaddition,[23] which
is one of the variations of click chemistry, has been exten-
sively used for the functionalisation of biomolecules,[41] in-
cluding the post-synthetic labelling of DNA.[25] We prepared
DMT-off ONs containing a single internal insertion of 2’-
deoxy-5-ethynyluridine (V),[26] (R)-1-O-(4-ethynylbenzyl)-
glycerol (Z)[28,42] or commercially available 2’-O-propargyl-
AHCTUNGTRENNUNG
uridine (Y)[27] (Table 1, ON4–ON9).
Previously, the palladium(0)-catalysed Sonogashira reac-
tion was performed on a solid support of ONs possessing 5-
iodopyrimidines,[38] 8-bromopurines[39] as well as 2- or 4-io-
dobenzyl- or 4-ethynylbenzylglycerols.[28,40] No side-reactions
were observed for native nucleobases with protecting
groups.[38] For the post-synthetic Sonogashira reaction using
automated DNA synthesis we prepared 5’-DMT-on ONs
ON1–ON3 containing a single internal insertion of one of
the following functionalised DNA building blocks: 2’-deoxy-
5-ethynyluridine (V; Scheme 1), 2’-deoxy-8-bromoguanosine
(W) or (R)-1-O-(2-iodobenzyl)glycerol (X; Scheme 4). The
phosphoramidites required for oligonucleotide synthesis
were either prepared, as in the case of 2’-deoxy-5-ethynyl-
Table 1. ONs before and after CuAAC reactions with porphyrin azides
24 or 25.[a]
Oligonucleotide (5’-3’)
m/z [Da]
Retention Conversion
time
[%][c]
[min][b]
calcd found
ON4
ON5
CCCCTTVCTTTTTT
CCCCTTYCTTTTTT
–
–
–
–
–
–
–
–
–
–
–
–
18.8
18.2
17.8
20.8
20.7
20.8
–
–
–
–
–
–
ON6 CCCCTTZTCTTTTTT
ON7
ON8
ON9
AGCTVGCTTGAG
AGCTYGCTTGAG
CTCAAGZCAAGCT
ON10 CCCCTT1CTTTTTT 4958.9 4956.0 41.6
ON11 CCCCTT2CTTTTTT 4972.9 4961.7 41.7
ON12 CCCCTT3CTTTTTT 4988.9 4979.8 42.0
ON13 CCCCTT4CTTTTTT 5002.9 4997.4 40.8
ON14 CCCCTT5TCTTTTTT 5216.9 5202.8 43.2
ON15 CCCCTT6TCTTTTTT 5231.0 5229.9 42.6
45
48
64
71
60
58
31
55
60
67
19
39
ON16
ON17
ON18
ON19
AGCT1GCTTGAG
AGCT2GCTTGAG
AGCT3GCTTGAG
AGCT4GCTTGAG
4497.8 4507.0 45.1
4511.9 4517.2 45.0
4527.8 4525.9 46.2
4541.8 4541.4 45.2
ON20 CTCAAG5CAAGCT 4693.9 4697.2 47.0
ON21 CTCAAG6CAAGCT 4707.9 4707.9 46.7
[a] Post-synthetic CuAAC reactions: DNA on CPG (0.33 mmol), azide 24
or 25 (7.67 mmol), CuSO4·5H2O (0.2 mmol in 5 mL H2O), sodium ascor-
bate (1.0 mmol in 20 mL H2O), DMSO (150 mL), shaking, RT, 3 days.
[b] Retention times Æ0.5 min, see the Experimental Section for HPLC
gradients. [c] The conversions of DNA were determined from the HPLC
peak integrals at 260 nm.
Scheme 4. Reagents and conditions: compound 12 or 15, [PdACTHUNTRGNE(UNG PPh3)4],
CuI, DMF, Et3N, 3 h.
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6231

V. V. Filichev et al.
CuAAC reactions were carried out by mixing nickel-con-
taining porphyrin azide 24 or 25 (7.67 mmol) and one of the
alkyne CPG-bound oligonucleotides ON4–ON9 (0.33 mmol)
in a micro-centrifuge vial followed by the addition of
DMSO (150 mL) to dissolve the azide. Freshly prepared
CuSO4·5H2O (0.2 mmol in 5 mL H2O) and sodium ascorbate
(1.0 mmol in 20 mL H2O) solutions were added, which result-
ed in the partial precipitation of the porphyrin azide. The re-
action vessels were then sealed under argon to avoid DNA
cleavage caused by copper ions in the presence of oxygen[43]
and shaken at RT for 3 days. After shaking, the CPG sup-
ports were repeatedly washed with CH2Cl2 to remove un-
reacted azide. The resulting red-coloured CPG support pro-
vided an indication of the progression of the reaction. The
unreacted azide was recovered in 80–90% yield for use in
future coupling reactions by washing the CH2Cl2 solution
containing the azide with H2O followed by drying over
MgSO4 and precipitation from CH2Cl2/MeOH. ONs were
then cleaved from the solid support (32% NH4OH) and pu-
rified by semi-preparative HPLC on a C18 column. Porphy-
rin-conjugated ONs showed appreciably higher retention
times than unmodified oligonucleotides (Table 1 and
Figure 1 in the Supporting Information) and ONs possessing
2’-O-propargyluridine (Y) showed superior conversions in
comparison with ONs containing 2’-deoxy-5-ethynyluridine
(V) and (R)-1-O-(4-ethynylbenzyl)glycerol (Z). Mixed
purine/pyrimidine sequences ON16–21 were found to have
lower conversions than pyrimidine sequences ON10–15. For
the incorporation of multiple porphyrins into DNA a proto-
col was developed[22] in which focused microwave irradia-
tion[44,45] (20 min at 708C) resulted in the complete conver-
sion of alkyne-DNA.
After HPLC purification, the collected fractions were
lyophilised, redissolved in 100 mL of H2O and precipitated
from LiClO4/acetone to give a deep-red pellet. The pellet
was dissolved in 100 mL of H2O to create stock solutions (ca.
300–1000 mm). Not all ON precipitates completely dissolved
in water (100 mL) and heating at 708C for several hours was
required to increase solubility. ONs were desalted using C18
zip-tips and then characterised by MALDI-TOF MS
(Table 1). The purity was checked by using 20% denaturing
polyacrylamide gel electrophoresis (PAGE; 7m urea), which
showed a single red band with significant retardation com-
pared with the wild-type ON (Figure 1). This is in contrast
to the work published by Fendt et al., which showed in-
creased mobility of ONs possessing a single porphyrin
moiety compared with the wild-type ON.[5]
Figure 1. Denaturing 20% PAGE (7m urea) of porphyrin–ONs stained
with Stains-All and destained in H2O. Porphyrin-modified ONs were red
before staining. Lane 1 is unmodified oligonucleotide ONwt (top) and a
colour marker (Bromophenol Blue, bottom), lane 2 is ON17, lane 3 is
ON16, lane 4 is ON19 and lane 5 is ON21.
between porphyrins (see Figure 2 of the Supporting Infor-
mation).
Further analysis revealed that the CD signals are pH-de-
pendent for the CT sequences (ON10–15) but not for the
mix-mers ON16–21. For example, the CT sequence ON15
showed virtually no CD signals in the porphyrin region at
pH 7.2, whereas a strong bisignate curve was observed at
pH 5.0 and 6.0 (Figure 2A). In addition, the CD intensity in
the UV region decreased with increasing pH and was blue-
shifted by 6 nm. This is a characteristic CD signature of i-
motifs,[47] which are formed by the base-pairing of hemi-pro-
tonated cytosine+ and cytosine to form duplex structures
that are zipped together in an anti-parallel orientation.[48]
This topology is enabled by the N3-protonation of cytosine,
which forms three hydrogen bonds with another cytosine.
Depending on the ionic strength, the pKa of cytosine in i-
motifs varies between 4.5 and 4.8,[49] which makes these
structures unstable at higher pH. It is important to note that
the CD signals of unmodified CT-ONwt (Table 2, see later)
did not show the characteristic pH dependency (Figure 2B),
which suggests that the porphyrin triggers the formation of
i-motifs for the sequence containing a run of only four cyto-
sines. An increase in the pH from 5.0 to 7.2 led to cytosine
deprotonation, i-motif unfolding and loss of porphyrin–por-
phyrin interactions. We assume that i-motif formation al-
lowed the orientation of porphyrins in such a manner that
porphyrin–porphyrin exciton coupling occurred through por-
phyrin stacking, as shown in the plausible i-tetraplex struc-
ture (Figure 2D). Native PAGE (pH 5.0) confirmed the ab-
Single-stranded porphyrin-ONs: Single-stranded ON10–
ON21 were prepared as 1.0 mm solutions in cacodylate
buffer at various pH (5.0, 6.0 and 7.2). The CD spectra of all
the porphyrin-modified ONs at pH 6.0 showed signals in the
region typical of DNA nucleobases at 248 (Àve ellipticity)
and 280 nm (+ve ellipticity, not shown).[46] Strong signals
arising from the absorbance of the porphyrin Soret band
were observed including the formation of bisignate curves,
which is an indication of the existence of exciton coupling
6232
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238

b-Pyrrolic-Modified Porphyrins
FULL PAPER
Figure 2. A) CD spectra of single-stranded oligopyrimidine ON15 at pH 5.0, 6.0 and 7.2. B) CD spectra of ONwt at pH 5.0, 6.0 and 7.2. C) CD thermal
melting of single-stranded oligopyrimidine ON15 at pH 5.0. D) Possible i-tetraplex formation. Buffers contain 20 mm sodium cacodylate, 100 mm NaCl
and 5 mm MgCl2.
sence of an i-motif for ONwt, whereas CT-ONs with por-
phyrins did not penetrate into the 20% native gel at pH 5.0
(gel not shown). This was also true with a decreased per-
centage of acrylamide gels (8 and 12%). We recently ob-
served[22] that short DNAs (12–20-mers) with more than two
porphyrins in their structures do not penetrate into native
and denaturing PAGEs, possibly as a result of aggregation
in the gel well. In the present case a significant retardation
of porphyrin DNAs compared with that observed in dena-
turing PAGE confirms the formation of a secondary struc-
ture (i-motif).
Thermal melting of the single-stranded ONs resulted in
the loss of the CD signal at around 430 nm and reduced
signal intensity at 280 nm (Figure 2C), which suggests the
melting of i-motifs (ON10–ON15) and the separation of ag-
gregates. We performed CD thermal denaturation of single-
stranded ON10, ON12 and ON14 at pH 5.0 and found that
irreversible melting occurred with mid-transition tempera-
ON16–21. Recently, intermolecular, end-to-end porphyrin–
porphyrin stacking between DNA duplexes and, conse-
quently, distinctive multi-signate CD signals in the porphyrin
Soret band region were observed for duplexes possessing a
porphyrin at the 5’-end at high ionic strength (150 mm NaCl
for copper–porphyrin, 450 mm for zinc–porphyrin).[19,21]
Clearly, porphyrin–porphyrin intermolecular interactions in
the DNA structures are strongly dependent upon the ON
sequence, the length, the porphyrin insertion and the ionic
strength. For instance, in the course of our study of purine-
pyrimidine 18-mer DNA possessing modification 4 in the
middle of the sequence,[22] we did not observe aggregate for-
mation for single-stranded ONs probably due to increased
electrostatic repulsion between negatively charged phos-
phates. For comparison, our CT sequence is 14-mer and our
mix-mer sequence studied here is 12-mer. Nevertheless, the
aggregation of the ONs described herein was found to be
slow after heating our samples at 908C for 5 min and thus it
should not affect the formation of duplexes and triplexes.
1
tures (T = ) of 51, 52 and 568C, respectively (Table 1 of the
2
Supporting Information). After melting, the porphyrinic CD
signal was observed only after incubation at 208C overnight
and with an intensity significantly lower than that of unheat-
ed samples. This means that the formation of porphyrinic i-
motif aggregates is slow at 1.0 mm and that aggregates pre-
sumably form in the concentrated stock solution (ca. 300–
1000 mm). To the best of our knowledge this is the first ob-
servation of ON aggregation/i-motif formation as a result of
internal incorporation of a single porphyrin in the sequence.
The pH-independent CD spectra observed for mix-mers
(see Figure 3 of the Supporting Information) suggest the for-
mation of aggregates rather than i-motifs in sequences
DNA duplexes and triplexes: The thermal stability of du-
plexes and triplexes containing the synthesised oligonucleo-
tides ON10–ON21 was assessed by UV/Vis thermal denatu-
ration experiments in the range 10–708C at 260 and 430 nm.
The melting temperatures (Tm) were determined as the
maxima of the first derivatives of the melting curves and
they are listed in Tables 2 and 3 (annealing temperatures are
listed in Tables 2 and 3 of the Supporting Information). Se-
quences possessing different porphyrin modifications were
studied in a parallel triplex with the duplex D1 and in anti-
parallel duplex with appropriate oligonucleotides ON22–24.
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6233

V. V. Filichev et al.
Table 2. Melting temperatures of porphyrin-containing parallel triplex and anti-parallel duplex.[a]
two orientations are possible as
a result of the long linker be-
tween the porphyrin and the
phosphate backbone (19 ꢂ). In-
tercalation, similar to that
shown in twisted intercalating
nucleic acids (TINA) configura-
tions,[44] was found to be of
higher energy (À20034 kJmolÀ1,
Figure 3B) than when the por-
phyrin penetrates the duplex
strands and aligns itself in the
minor groove (À20086 kJmolÀ1,
Figure 3C).
Sequence (5’-3’)
Parallel triplex[b] 3’-
CTGCCCCTTTCTTTTTT/5’-GACGGG-
GAAAGAAAAAA (D1)
Anti-parallel duplex[c] 3’-
GGGGAAAGAAAAAA
(ON22)
pH 5.0
pH 6.0
pH 6.0
pH 7.2
ONwt CCCCTTTCTTTTTT
ON10 CCCCTT1CTTTTTT
ON11 CCCCTT2CTTTTTT
ON12 CCCCTT3CTTTTTT
ON13 CCCCTT4CTTTTTT
ON14 CCCCTT5TCTTTTTT
ON15 CCCCTT6TCTTTTTT
54.0[d]
27.0
48.0
48.0
39.0 (39.0)
40.0 (39.5)
54.5[d] (53.5)
55.0[d] (54.0)
55.0[d] (53.0)
55.0[d] (53.5)
39.0 (37.0)
30.0 (37.7)
34.5 (29.3)
34.5 (32.7)
38.9 (37.0)
36.3 (36.8)
31.0 (30.0)
32.5 (29.0)
35.0 (28.5)
41.5 (43.8)
38.7 (24.0)
33.3 (31.0)
30.0 (30.3)
33.2 (33.7)
33.2 (33.7)
41.7 (43.0)
34.6 (37.0)
34.3 (34.6)
[a] Tm [8C] data for parallel triplex and anti-parallel duplex melting, taken from the UV/Vis melting curves at
260 and 430 nm (shown in parentheses). [b] 1.5 mm of ON10–15 and ONwt and 1.0 mm of each strand of
dsDNA (D1) in 20 mm sodium cacodylate, 100 mm NaCl and 5 mm MgCl2, pH 5.0 and 6.0. [c] 1.0 mm of each
strand in 20 mm sodium cacodylate, 100 mm NaCl and 5 mm MgCl2, pH 6.0 and 7.2. [d] Third strand and
duplex melting overlaid.
As can be seen from the Tm data in Table 2, the internal
insertion of a porphyrin resulted in increased Tm values
(DTm =3.0–12.08C) of the Hoogsteen-type triplexes relative
to the wild-type triplex ONwt/D1 at pH 6.0. The TFO
strands ON10 and ON14 containing aromatic porphyrins
(modifications 1 and 5, Scheme 1) resulted in slightly greater
stabilisation of the triplex at pH 6.0 compared with ON11
and ON15 containing the aliphatic-linked porphyrins (modi-
fications 2 and 6). Note that the pH dependence of triplex
thermal stability is less pronounced for triplexes in which
the porphyrin is linked to the 5-position of uridine (ON10/
D1 and ON11/D1 versus ON12–ON15/D1, pH 5.0 and 6.0).
This can be explained by molecular modelling (AMBER*
force field,[50] Figure 3). The porphyrins in modifications 1
and 2 are located in the major groove of the duplex, occupy-
ing the space between the TFO and the homopyrimidine
strand of the duplex D1 (Figure 3A). This means that the
porphyrin, especially in the case of aliphatic modification 2,
is located close to the nucleobases. This hydrophobic effect
alters the hydration of the triplex and might have an influ-
ence on the pKa values of cytidines nearby, thus changing
the response of the triplex to pH. In contrast, the porphyrins
in modifications 3 and 4 are located away from the nucleo-
bases on the edge of the homopurine strand of the duplex
and the TFO with very little influence on the hydration of
the major groove of the DNA (Figure 3B). Aliphatic por-
phyrins provide greater flexibility in comparison with aro-
matic porphyrins, which results in closer interactions be-
Figure 3. Representations of the lowest-energy structures of triplexes
ON10/D1 (A), ON12/D1 (B) and ON14/D1 (C, D) possessing aromatic
porphyrins 1, 3 and 5, respectively. Two possible porphyrin orientations
are shown for the triplex ON14/D1: C) The porphyrin penetrating
through the duplex D1. D) The higher-energy structure in which the por-
phyrin intercalates between the bases of duplex D1. E) 1,2,3-Triazole-
linked benzyl moiety incorporated into TFO as a bulge.[44]
tween the porphyrin and the triplex (see Figure 5 of the
Supporting Information). We performed thermal stability
studies on triplexes having a single mis-match with ON10–
13 and found that no triplexes are formed at pH 5.0 (data
not shown). This means that there is no intercalation of the
porphyrin in triplexes ON10–13/D1, in accord with results
of the molecular modelling. In the case of ON14 and ON15,
in which the porphyrin is incorporated as a bulged insertion
into the triplex strands, the porphyrin may either penetrate
through the DNA duplex (Figure 3C) or intercalate be-
tween bases in the duplex such that the meso-phenyl rings
are located in the grooves of the triplex (Figure 3D). These
This difference in energy is most likely a result of the de-
stabilising bulge in the phosphate backbone of the TFO
strand, which is required for porphyrin intercalation be-
tween the bases of the duplex strands. Further thermal stabi-
6234
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238

b-Pyrrolic-Modified Porphyrins
FULL PAPER
lisation might be possible by reducing the length of the
linker, which reduces the bulge in the backbone of the TFO
strand.
ON14,15/ON22 over ON10,11/ON22. As the porphyrin is a
hydrophobic moiety it could be expected that the destabili-
sation observed is a result of a significantly changed hydra-
tion of the duplex in comparison with the wild-type duplex.
Significant differences were found in the melting tempera-
tures of duplexes containing ON12 and ON13 in which the
porphyrin is positioned in the minor groove. Duplexes con-
taining the aromatic porphyrin 3 (ON12) show DTm values
of À13.0 to À14.88C, respectivelyat pH 6.0 and 7.2, similar
to the duplexes ON10,11,14,15/ON22, whereas duplex
ON13/ON22, which contains the aliphatic porphyrin 4, has a
DTm of À6.3 to À6.58C. Molecular modelling shows that the
aromatic (ON12, Figure 4B) and aliphatic porphyrins
(ON13, Figure 4E) follow the minor groove in a similar
manner. However, the aliphatic porphyrin is less destabilis-
ing, presumably due to rotational flexibility around the
extra sp3 carbon which results in a better fit within the
minor groove in comparison with the aromatic porphyrin.
Duplexes containing mixed purine/pyrimidine strands
(Table 3) show a destabilisation trend similar to the porphy-
rin-modified CT sequences. Thus, significant thermal desta-
As expected, the Tm values of the triplexes of ON12–15/
D1 increased at pH 5.0 due to increased cytosine protona-
tion. Surprisingly, the Tm values of ON10 and ON11 were
almost unchanged. No triplex formation was observed
above 108C at pH 7.2. The importance of the porphyrin in
the stability of the triplex can be emphasised when we com-
pare the previously reported Tm value for the bulged inser-
tion of the 1,2,3-triazole-linked benzyl moiety (Figure 3E)
in the TFO strand.[44] Under identical conditions and by
using the sequence equivalent to ON15, destabilisation of
the resulting triplex by more than 238C was observed at
pH 6.0.
To confirm triplex formation, the CD spectra of the tri-
plexes were recorded at pH 5.0 and 6.0. A strong negative
band at 209 nm was observed (see Figure 4 of the Support-
ing Information), which confirms the formation of a tri-
plex[51] in porphyrin-modified sequences.
It was observed (Table 2) that a single internal incorpora-
tion of a porphyrin in an anti-parallel duplex (ON10–15/
ON22) resulted in the thermal destabilisation of the duplex
compared with the unmodified duplex. Variation of the de-
stabilisation depended on the porphyrin incorporated and
its location in the duplex. We found that duplexes involving
ON10,11,14,15 showed a significant drop in thermal stability
of 13.4–18.08C per modification at pH 7.2. Molecular model-
ling of these duplexes (Figure 4) suggest that the porphyrins
have very little interaction with the nucleobases and are
positioned in the major groove. As a result of the linker
length, porphyrins 5 and 6 protrude through the duplex into
the major groove, leaving just the phenyl and triazole moiet-
ies located between the bases in the duplex core. This inter-
calation may account for the slight duplex stabilisation of
Table 3. Melting temperatures of anti-parallel duplexes containing mixed
purine/pyrimidine strands.[a]
Sequence (5’-3’)
Anti-parallel duplex Anti-parallel duplex
3’-TCGAAC-
GAACTC (ON23)
3’-GAGTTCGTTC-
GA (ON24)
pH 6.0
pH 7.2
pH 6.0
pH 7.2
ONwtm AGCTTGCTTGAG
50.0
35.0
(32.0)
34.0
(33.5)
36.0
(36.0)
42.0
50.0
35.0
ACHTUNGTRNE(NUNG 34.0)
–
–
–
–
ON16
ON17
ON18
ON19
AGCT1GCTTGAG
AGCT2GCTTGAG
AGCT3GCTTGAG
AGCT4GCTTGAG
32.0
–
–
–
–
–
–
(NVT)[b]
34.0
ACHTUNGTRNE(NUNG 35.0)
43.0
(39.0)
–
(NVT)[b]
–
ON20 CTCAAG5CAAGCT
ON21 CTCAAG6CAAGCT
32.0
(33.0)
29.0
33.3
(NVT)[b]
27.9
–
–
(28.0)
ACHTUNGTRENNUNG(29.1)
[a] Tm [8C] data for anti-parallel duplex melting, taken from the UV/Vis melt-
ing curves at 260 and 430 nm (shown in parentheses). 1.0 mm of each strand in
20 mm sodium cacodylate, 100 mm NaCl and 5 mm MgCl2, pH 6.0 and 7.2.
[b] NVT=no visible transition.
bilisation was found for duplexes containing bulge insertions
of porphyrin-glycerol moieties (ON20 and ON21 with
ON24). Minor groove modification with the aliphatic por-
phyrin 4 resulted in the least destabilised duplex (ON19/
ON23). In addition, the melting profiles (see Tables 2 and 3
of the Supporting Information) show virtually no hysteresis
at either 260 or 430 nm for all duplexes, which indicates that
the kinetics of both the denaturing and annealing processes
are fast in comparison with the temperature gradient.
The CD spectra of duplexes containing ON10–21 show a
negative band at around 245–250 nm and a positive band at
274–280 nm (see Figure 6 of the Supporting Information),
which clearly suggests that the modified duplexes retain
Figure 4. Representations of the lowest-energy structures of the duplexes
formed by ON10 (A), ON12 (B), ON14 (C), ON11 (D), ON13 (E) and
ON15 (F) with ON22.
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6235

V. V. Filichev et al.
their overall B-form double-helix structure. A CD signal in
the porphyrin Soret band region was also observed for these
duplexes either as a weak band or as a bisignate curve. In
contrast to the CD spectra of single-stranded CT sequences,
signals of duplexes in the porphyrin region returned to the
pre-melted level almost instantaneously after samples were
heated and cooled. Moreover, marginal changes in the CD
profiles were observed by using lower or higher NaCl con-
centrations (50 or 500 mm, respectively, data not shown).
Thus, duplex aggregation triggered by porphyrins, which
should occur over longer periods of time and at high salt
concentrations, was thought unlikely to happen. The pres-
ence of these CD signals, especially for constructs possessing
a methylene group between a porphyrin and a nucleotide,
led to the conclusion that the porphyrin interacts with the
chiral environment of DNA.
positioned in the minor groove of the duplex when attached
to the 2’-O-position of the uridine. In contrast, porphyrin at-
tachment at the 5-position of uridine or its use as an interca-
lating moiety led to duplexes with porphyrin in the major
groove with no or very few interactions with nucleobases.
Presumably, the additional flexibility provided by two meth-
ylene linkages in compound 4 helps the porphyrin to fit
better within the minor groove of the duplex in comparison
with the aromatic porphyrin 3. These findings highlight the
difference between the positioning of porphyrins in different
DNA environments and thus they can be used in the con-
struction of DNA-based photonic devices.
Experimental Section
Solid-phase synthesis of oligonucleotides and post-synthetic CuAAC:
DMT-off oligodeoxynucleotides were synthesised by using MerMade 4
Automated DNA Synthesiser from BioAutomation Corporation on a
1.0 mmol scale on 1000 ꢂ CPG supports using 4,5-dicyanoimidazole as an
activator and 0.075m solutions of the corresponding phosphoramidites of
V, Y and Z in dry MeCN with an increased coupling time (2 min). After
DNA synthesis, DMT-off ON4–ON9 on CPG (0.33 mmol) containing V,
Y or Z were removed from their corresponding columns and added to a
microcentrifuge vial (1.5 mL) followed by the appropriate azide
(7.67 mmol, 23 equiv) in degassed DMSO (150 mL). Freshly prepared
CuSO4·5H2O (0.2 mmol, 0.6 equiv, 5 mL of a 40 mm solution in degassed
H2O) and sodium ascorbate (1.0 mmol, 3 equiv, 20 mL of a 50 mm solution
in degassed H2O) were added. The reaction mixture was shaken over
argon in darkness for 3 days. CH2Cl2 (1 mL) was added to the reaction
mixture and the CPGs were centrifuged (14500 rpm for 1 min). The sol-
vents were removed and washing was repeated until the supernatant no
longer showed any colour (see recovery of porphyrin azides). The red
CPG was then washed with H2O (1.5 mL) to remove any remaining inor-
ganic salts. Residual solvent was removed under reduced pressure and
the obtained DMT-off oligonucleotides bound to CPG supports were
treated with 32% aq. NH4OH (1 mL) at RT for 2 h and then at 558C
overnight. The porphyrin-functionalised DMT-off ONs were purified by
using a Waters 600 HPLC apparatus fitted with a Waters 2487 dual l ab-
sorbance detector (260 and 427 nm) with a reversed-phase semi-prepara-
tive Econosil C18 (10 mm, 10ꢄ250 mm) column. Buffer A (0.05m triethy-
lammonium acetate in H2O, pH 7.0) and buffer B (75% CH3CN in H2O),
flow 2.5 mLminÀ1. Gradients: 2 min 100% A, linear gradient to 70% B
in 38 min, linear gradient to 100% B in 7 min, 100% B in 3 min and then
100% A in 10 min. After purification, the corresponding fractions were
lyophilised, dissolved in H2O (100 mL, heating to 708C for 1 h was re-
quired to dissolve some ONs), 0.01m lithium perchlorate in acetone
(1.6 mL) was added and the ONs were stored at À108C for 1 h. The pre-
cipitated ON pellet was centrifuged (15000 rpm for 30 min), the superna-
tant removed and the pellet was washed with acetone (30 mL). The mo-
lecular weights of the oligonucleotides were obtained by using a Bruker
Daltonics Autoflex MALDI-TOF spectrometer in the negative mode
using either 2’,4’,6’-trihydroxyacetophenone, 3-hydroxypicolinic acid or 6-
azathiothymine matrices and dibasic ammonium citrate or 1H-imidazole
as co-matrix. Oligonucleotides were desalted by using C18 ZipTips (Milli-
pore) prior to loading on the MALDI plate. The purity was checked by
using denaturing 20% PAGE (7m urea), which showed a single red band
with significant retardation relative to the wild-type oligonucleotide.
Conclusion
We have investigated two routes of attachment of b-pyrrol-
ic-substituted porphyrins to nucleosides and DNA. Several
2’-deoxyuridine derivatives were synthesised by either Sono-
gashira or CuAAC reactions. A higher degree of conjuga-
tion between uracil and the porphyrin core connected
through a 1,4-diethynylbenzene moiety (19) was observed in
the UV/Vis spectrum in comparison with the previously
studied meso-functionalised porphyrin (20) linked at the 5-
position of 2’-deoxyuridine. However, due to a low yields of
porphyrin-nucleosides by the Sonogashira reaction we
switched our focus to CuAAC coupling, which led to nu-
cleoside and DNA conjugates modified with azidostyryl and
methylazidostyryl-substituted nickel(II) porphyrins. We
screened the effect of various single porphyrin modifications
on the structures and thermal stability of single-stranded
ONs and duplex and triplex DNAs. Three types of DNA
building blocks with terminal alkynes were considered for
CuAAC coupling to DNA: 2’-deoxy-5-ethynyluridine, 2’-O-
propargylACHTUNGTRENNUNGuridine and (R)-4-ethynylphenylmethylglycerol.
Single-stranded ONs containing internal porphyrin modifi-
cations formed porphyrin-driven i-motif structures in pyri-
midine sequences having a single terminal run of four cyto-
sines. Porphyrin–porphyrin intermolecular interactions were
also detected for purine/pyrimidine sequences. In both cases,
the formation of aggregates/i-motifs was slow at room tem-
perature after heating samples at 958C for 5 min. Thermal
stability studies were performed on Hoogsteen-type triplex-
es containing porphyrin-modified TFO strands and it was
found that porphyrin modifications generally stabilise the
triplex in the DTm range of +3 to +128C. This shows that
2’-deoxy-5-ethynyluridine and 2’-O-propargyluridine are
promising as ꢃclickableꢀ nucleotides for labelling with organ-
ic chromophores in internal positions of TFOs. Duplex ther-
mal stability studies revealed that the aliphatic porphyrin 25
internally connected to the 2’-position of uridine destabilises
duplexes to a lesser extent than other modifications. Ac-
cording to molecular modelling studies, porphyrins were
Recovery of azides: H2O (50 mL) was added to the combined CH2Cl2
washings containing the reacted porphyrin azide and the resulting solu-
tion was vigorously stirred for 1 h. The organic layer was extracted into
CH2Cl2 (2ꢄ50 mL), dried over MgSO4, filtered and the porphyrin was
precipitated from CH2Cl2/MeOH. The desired product was collected by
filtration to give a red solid (approximately 80–90% recovery).
6236
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238

b-Pyrrolic-Modified Porphyrins
FULL PAPER
UV/Vis spectroscopic melting temperature measurements: Melting tem-
Acknowledgements
peratures for duplexes and triplexes were measured with
a CARY
100Bio UV/Vis spectrophotometer using a 2ꢄ6 Multicell block with Pelt-
ier temperature controller. The triplexes were formed by first mixing the
two strands of the Watson–Crick duplex, each at a concentration of
1.0 mm, in the corresponding buffer solution followed by the addition of
the third TFO strand at a concentration of 1.5 mm (total volume 1 mL).
The solutions were then heated at 908C for 15 min, cooled and incubated
at 108C for 30 min. The duplexes were formed by mixing the two strands,
each at a concentration of 1.0 mm, in the appropriate buffer (total volume
1 mL). The solutions were heated at 908C for 15 min, cooled and incubat-
ed at 108C for 30 min. The melting temperatures were determined as the
maxima of the first derivative plots of the melting curves obtained by
measuring the absorbance at 260 and 430 nm against increasing tempera-
ture (10–708C, 1.08CminÀ1 for duplexes and 10–708C, 0.58CminÀ1 for
triplexes). All melting temperatures were an average of two denaturing/
annealing cycles. Extinction coefficients for porphyrin-modified oligonu-
cleotides were calculated by using the extinction coefficients of each nu-
cleoside at 260 nm. Extinction coefficients for porphyrin modified-nu-
cleosides at 260 nm: 1, 14800; 2, 14800; 3, 15900; 4, 15900; 5, 9000; 6,
This work was financially supported by Marsden grants administrated by
the Royal Society of New Zealand (grants nos. MAU0407 and
MAU0704).
[1] a) V. V. Filichev, E. B. Pedersen, Wiley Encycl. Chem. Biol. 2009, 1,
493–524; b) R. Varghese, H. A. Wagenknecht, Chem. Commun.
2009, 2615–2624; c) H. A. Wagenknecht, Angew. Chem. 2009, 121,
2878–2881; Angew. Chem. Int. Ed. 2009, 48, 2838–2841; d) V. L.
Malinovskii, D. Wenger, R. Haner, Chem. Soc. Rev. 2010, 39, 410–
422.
[2] a) W. M. Campbell, K. W. Jolley, P. Wagner, K. Wagner, P. J. Walsh,
K. C. Gordon, L. Schmidt-Mende, M. K. Nazeeruddin, Q. Wang, M.
Gratzel, D. L. Officer, J. Phys. Chem. C 2007, 111, 11760–11762;
b) W. M. Campbell, A. K. Burrell, D. L. Officer, K. W. Jolley, Coord.
Chem. Rev. 2004, 248, 1363–1379.
[3] L. M. Moreira, F. V. dos Santos, J. P. Lyon, M. Maftoum-Costa, C.
Pacheco-Soares, N. S. da Silva, Aust. J. Chem. 2008, 61, 741–754.
[4] I. Bouamaied, L. A. Fendt, D. Haussinger, M. Wiesner, S. Thoni, N.
Amiot, E. Stulz, Nucleosides Nucleotides Nucleic Acids 2007, 26,
1533–1538.
[5] L. A. Fendt, I. Bouamaied, S. Thoni, N. Amiot, E. Stulz, J. Am.
Chem. Soc. 2007, 129, 15319–15329.
[6] I. Bouamaied, T. Nguyen, T. Ruhl, E. Stulz, Org. Biomol. Chem.
2008, 6, 3888–3891.
[7] T. Nguyen, A. Brewer, E. Stulz, Angew. Chem. 2009, 121, 2008–
2011; Angew. Chem. Int. Ed. 2009, 48, 1974–1977.
[8] M. Endo, M. Fujitsuka, T. Majima, Tetrahedron 2008, 64, 1839–
1846.
[9] a) K. Borjesson, J. Tumpane, T. Ljungdahl, L. M. Wilhelmsson, B.
Norden, T. Brown, J. Martensson, B. Albinsson, J. Am. Chem. Soc.
2009, 131, 2831–2839; b) K. Borjesson, J. Wiberg, A. H. El-Sagheer,
T. Ljungdahl, J. Martensson, T. Brown, B. Norden, B. Albinsson,
Acs Nano 2010, 4, 5037–5046.
[10] H. Morales-Rojas, E. T. Kool, Org. Lett. 2002, 4, 4377–4380.
[11] S. Sitaula, S. M. Reed, Bioorg. Med. Chem. Lett. 2008, 18, 850–855.
[12] M. Balaz, A. E. Holmes, M. Benedetti, G. Proni, N. Berova, Bioorg.
Med. Chem. 2005, 13, 2413–2421.
[13] M. Balaz, K. Bitsch-Jensen, A. Mammana, G. A. Ellestad, K. Naka-
nishi, N. Berova, Pure Appl. Chem. 2007, 79, 801–809.
[14] A. Mammana, T. Asakawa, K. Bitsch-Jensen, A. Wolfe, S. Chaturan-
tabut, Y. Otani, X. X. Li, Z. M. Li, K. Nakanishi, M. Balaz, G. A. El-
lestad, N. Berova, Bioorg. Med. Chem. 2008, 16, 6544–6551.
[15] M. Balaz, A. E. Holmes, M. Benedetti, P. C. Rodriguez, N. Berova,
K. Nakanishi, G. Proni, J. Am. Chem. Soc. 2005, 127, 4172–4173.
[16] a) M. Balaz, B. C. Li, S. Jockusch, G. A. Ellestad, N. Berova, Angew.
Chem. 2006, 118, 3610–3613; Angew. Chem. Int. Ed. 2006, 45, 3530-
3533; b) M. Balaz, B. C. Li, J. D. Steinkruger, G. A. Ellestad, K. Na-
kanishi, N. Berova, Org. Biomol. Chem. 2006, 4, 1865–1867; c) M.
Balaz, J. D. Steinkruger, G. A. Ellestad, N. Berova, Org. Lett. 2005,
7, 5613–5616.
9000 LmolÀ1 cmÀ1
.
Circular dichroism measurements: CD spectra were recorded using an
Applied Photophysics Chirascan CD spectrometer (150 W Xe arc) with a
Quantum Northwest TC125 temperature controller. Solutions containing
1.0 mm of each strand were used for CD spectroscopy. An average of 10
scans was recorded (1 nm intervals, 240 nmminÀ1, 1 cm pathlength), base-
lined against the appropriate buffer solution and then smoothed. Data
were recorded in mdeg and converted into De. CD spectroscopic melting
temperature measurements of i-motifs were preformed by using solutions
containing the appropriate strand at a concentration of 1.0 mm in pH 5.0
buffer. Melting temperatures were determined from the maxima of the
first derivative curves measured at 283 nm against increasing temperature
(25–908C at 58C intervals with 300 s equilibration time).
Molecular modelling: Molecular modelling calculations and duplex con-
struction were performed by using MacroModel v9.8 from Schrçdinger.
All calculations were performed by using the AMBER* force field[50] and
the GB/SA water model.[52] The 8-mer duplexes containing unmetallated
tetraphenylporphyrin-modified nucleotides were generated from a B-
type DNA–DNA duplex by using Maestro v9.8 from Schrçdinger. Paral-
lel triplexes were constructed by consecutive superimposition of CGC
and TAT triplets, which were generated with Insight II v9.72 from MSI[53]
and transported into MacroModel. Bulged insertions 5 and 6 were con-
structed in several steps. The unmodified duplex or triplex was construct-
ed and the appropriate strand was disconnected by removal of the phos-
phate group at the location of the bulged insertion. Afterwards, the
bulged insertion up to and including the triazole ring was constructed
and placed between bases in the structure. The phosphate backbone was
reconnected and minimisation was then performed, creating enough
space for the bulged insertion. The appropriate porphyrin was then
linked the triazole ring. Constraints ensured the planarity of the porphyr-
ins and were based on the metal-complexed porphyrin available in the
À
À
Maestro software (distances: N1 N3, N2 N4 4.132 ꢂ; force constant
100 kJmol-À1 ꢂÀ2; torsion angles: N1 C2 C3 C4, N2 C6 C7 C8, N3
À
À
À
À
À
À
À
C10 C11 C12, N4 C14 C15 C16 0.08; force constant 100 kJmolÀ1 ꢂÀ2).
Stochastic dynamics calculations generating 250 structures were per-
formed by using an extended cut-off potential with a SHAKE algorithm
to constrain bonds to hydrogen atoms. The simulation temperature was
300 K, the simulation time 500 ps and the equilibration time 150 ps. All
250 structures were minimised by using the Polak–Ribiere Conjugate
Gradient (PRCG) method with a convergence threshold of 0.05 kJmolÀ1
and examined with XCluster from Schrçdinger to find representative
low-energy structures.
À
À
À
À
À
[17] M. Endo, M. Fujitsuka, T. Majima, J. Org. Chem. 2008, 73, 1106–
1112.
[18] K. Berlin, R. K. Jain, M. D. Simon, C. Richert, J. Org. Chem. 1998,
63, 1527–1535.
[19] A. Onoda, M. Igarashi, S. Naganawa, K. Sasaki, S. Ariyasu, T. Ya-
mamura, Bull. Chem. Soc. Jpn. 2009, 82, 1280–1286.
[20] H. D. Li, O. S. Fedorova, W. R. Trumble, T. R. Fletcher, L. Czucha-
jowski, Bioconjugate Chem. 1997, 8, 49–56.
[21] A. Mammana, G. Pescitelli, T. Asakawa, S. Jockusch, A. G. Petrovic,
R. R. Monaco, R. Purrello, N. J. Turro, K. Nakanishi, G. A. Ellestad,
M. Balaz, N. Berova, Chem. Eur. J. 2009, 15, 11853–11866.
[22] A. W. I. Stephenson, N. Bomholt, A. Partridge, V. V. Filichev, Chem-
BioChem 2010, 11, 1833–1839.
[23] a) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless,
Angew. Chem. 2002, 114, 2708–2711; Angew. Chem. Int. Ed. 2002,
Chem. Eur. J. 2011, 17, 6227 – 6238
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6237

V. V. Filichev et al.
41, 2596–2599; b) C. W. Tornoe, C. Christensen, M. Meldal, J. Org.
Chem. 2002, 67, 3057–3064.
[40] a) V. V. Filichev, E. B. Pedersen, J. Am. Chem. Soc. 2005, 127,
14849–14858; b) V. V. Filichev, I. V. Astakhova, A. D. Malakhov,
V. A. Korshun, E. B. Pedersen, Chem. Eur. J. 2008, 14, 9968–9980.
[41] V. Hong, S. I. Presolski, C. Ma, M. G. Finn, Angew. Chem. 2009, 121,
10063–10067; Angew. Chem. Int. Ed. 2009, 48, 9879–9883.
[42] D. Globisch, N. Bomholt, V. V. Filichev, E. B. Pedersen, Helv. Chim.
Acta 2008, 91, 805–818.
[24] S. H. Weisbrod, A. Marx, Chem. Commun. 2008, 5675–5685.
[25] a) A. H. El-Sagheer, T. Brown, Chem. Soc. Rev. 2010, 39, 1388–
1405; b) P. M. E. Gramlich, C. T. Wirges, A. Manetto, T. Carell,
Angew. Chem. 2008, 120, 8478–8487; Angew. Chem. Int. Ed. 2008,
47, 8350–8358.
[26] D. Graham, J. A. Parkinson, T. Brown, J. Chem. Soc. Perkin Trans. 1
1998, 1131–1138.
[27] M. Grotli, M. Douglas, R. Eritja, B. S. Sproat, Tetrahedron 1998, 54,
5899–5914.
[28] I. Gꢅci, V. V. Filichev, E. B. Pedersen, Bioconjugate Chem. 2006, 17,
950–957.
[43] R. L. Weller, S. R. Rajski, Org. Lett. 2005, 7, 2141–2144.
[44] I. Gꢅci, V. V. Filichev, E. B. Pedersen, Chem. Eur. J. 2007, 13, 6379–
6386.
[45] C. Bouillon, A. Meyer, S. Vidal, A. Jochum, Y. Chevolot, J. P. Cloar-
ec, J. P. Praly, J. J. Vasseur, F. Morvan, J. Org. Chem. 2006, 71, 4700–
4702.
[29] M. Nath, J. C. Huffman, J. M. Zaleski, J. Am. Chem. Soc. 2003, 125,
11484–11485.
[46] J. Kypr, I. Kejnovska, D. Renciuk, M. Vorlickova, Nucleic Acids Res.
2009, 37, 1713–1725.
[30] E. Schwartz, S. Le Gac, J. Cornelissen, R. J. M. Nolte, A. E. Rowan,
Chem. Soc. Rev. 2010, 39, 1576–1599.
[31] E. E. Bonfantini, A. K. Burrell, W. M. Campbell, M. J. Crossley, J. J.
Gosper, M. M. Harding, D. L. Officer, D. C. W. Reid, J. Porphyrins
Phthalocyanines 2002, 6, 708–719.
[32] H. Firouzabadi, N. Iranpoor, S. Sobhani, Tetrahedron 2004, 60, 203–
210.
[33] A. W. I. Stephenson, P. Wagner, A. C. Partridge, K. W. Jolley, V. V.
Filichev, D. L. Officer, Tetrahedron Lett. 2008, 49, 5632–5635.
[34] E. E. Bonfantini, D. L. Officer, Tetrahedron Lett. 1993, 34, 8531–
8534.
[47] a) G. Manzini, N. Yathindra, L. E. Xodo, Nucleic Acids Res. 1994,
22, 4634–4640; b) V. Mathur, A. Verma, S. Maiti, S. Chowdhury,
Biochem. Biophys. Res. Commun. 2004, 320, 1220–1227.
[48] a) K. Gehring, J. L. Leroy, M. Gueron, Nature 1993, 363, 561–565;
b) J. L. Leroy, K. Gehring, A. Kettani, M. Gueron, Biochemistry
1993, 32, 6019–6031.
[49] J. L. Mergny, L. Lacroix, X. G. Han, J. L. Leroy, C. Helene, J. Am.
Chem. Soc. 1995, 117, 8887–8898.
[50] a) MacroModel, version 9.8, Schrodinger, LLC, New York, NY,
2010. b) S. J. Weiner, P. A. Kollman, D. T. Nguyen, D. A. Case, J.
Comput. Chem. 1986, 7, 230–252; c) S. J. Weiner, P. A. Kollman,
D. A. Case, U. C. Singh, C. Ghio, G. Alagona, S. Profeta, P. Weiner,
J. Am. Chem. Soc. 1984, 106, 765–784.
[35] M. Sꢅverac, L. Le Pleux, A. Scarpaci, E. Blart, F. Odobel, Tetrahe-
dron Lett. 2007, 48, 6518–6522.
[36] D. M. Shen, C. Liu, Q. Y. Chen, Eur. J. Org. Chem. 2007, 1419–
1422.
[37] J. M. Barbe, G. Canard, S. Brandes, R. Guilard, Eur. J. Org. Chem.
2005, 4601–4611.
[51] G. Manzini, L. E. Xodo, D. Gasparotto, F. Quadrifoglio, G. A. Van-
dermarel, J. H. Vanboom, J. Mol. Biol. 1990, 213, 833–843.
[52] W. C. Still, A. Tempcyk, R.C. Hawley, T. Hendrickson, J. Am.
Chem. Soc. 1990, 112, 6127-6129.
[38] a) A. E. Beilstein, M. W. Grinstaff, Chem. Commun. 2000, 509–510;
b) S. I. Khan, M. W. Grinstaff, J. Am. Chem. Soc. 1999, 121, 4704–
4705.
[53] Insight II, version 9.72, Accelrys Inc., San Diego, CA, 2005.
[39] a) M. Rist, N. Amann, H. A. Wagenknecht, Eur. J. Org. Chem. 2003,
2498–2504; b) E. Mayer, L. Valis, C. Wagner, M. Rist, N. Amann,
H. A. Wagenknecht, ChemBioChem 2004, 5, 865–868.
Received: November 8, 2010
Revised: February 1, 2011
Published online: April 18, 2011
6238
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6227 – 6238